dissenting.
I respectfully dissent. I do not understand how to apply the harm analysis the Court of Criminal Appeals tells us to use in Rose v. State, 752 S.W.2d 529 (Tex.Crim.App.1987) (op. on reh’g).
Since issuing Rose, the Court of Criminal Appeals has remanded 390 cases to the courts of appeals to decide whether the parole instruction1 caused harm. The remand of these cases is one of the most important issues in our criminal justice system today.
This Court and others have struggled over Rose and how we are to decide the cases on remand. Rose offers little help. And, because the Court of Criminal Appeals has refused to review any of the decisions applying the harm analysis to the parole instruction, it does not seem that we will get any guidance from that court.
The decisions of the courts of appeals are in hopeless conflict.2 The Court of Criminal Appeals has refused petitions in cases that have the same statements of the law *484with different results. Until the Court of Criminal Appeals grants petitions to review some of these cases, we will continue to operate without any instructions on this matter.
Without guidance other than Rose, we are forced to try to understand what the Court of Criminal Appeal meant when it issued Rose. Reading Rose carefully does not improve one’s understanding of it.
I. THE CONFUSION IN ROSE
There are two distinct parts to the Rose decision: Rose I, the first opinion holding the parole instruction was unconstitutional, Rose I, 752 S.W.2d 529-37; and Rose II, the opinion on rehearing stating the harm analysis we are to apply on remand. Rose II, 752 S.W.2d at 752-55. Rose I and Rose II seem irreconcilable. Below is a chart summarizing some of the inconsistencies.
[[Image here]]
If it is of “no constitutional consequence” the jury is told to ignore parole, how can the Court presume the jury followed the instructions when determining harm?
If the instruction “directly instructs” the jury to consider parole, how can the Court rely on the judge’s instruction that parole was not their concern?
*485If “experience teaches us” that the jury probably considered parole, how can the Court be ignorant of the process the jury underwent in assessing punishment?
If it can “never be properly discovered” if the jury considered parole, how is it we are to presume the jury followed the instructions to disregard it?
If the risk that punishment will be based on extraneous considerations is “intolerable,” how could the facts ever militate in favor of a harsh sentence?
If the parole instruction violates due process (which means the trial was fundamentally unfair), how is it the error on sentencing was harmless? See Chief Justice McCormick’s dissent in Rose II. 752 S.W.2d at 559.
In spite of these unanswered questions, I will try to determine if we can say, beyond a reasonable doubt, the parole instruction did not contribute to the punishment.
II. THE ROSE II GUIDELINES
Rose II gave us three factors to consider in making the harm analysis: (1) whether the trial court gave the curative instruction, (2) the heinous nature of the crime, and (3) the defendant’s criminal record. Rose II, 752 S.W.2d at 554.
A. THE INSTRUCTIONS TO DISREGARD.
There are two instructions in this charge to disregard parole: one, in the statutory language of the parole instruction in paragraph 9; and two, the curative instruction 3 in paragraph 11. The Rose II Court was
particularly impressed with the curative instruction because it was “the judge’s last word on the subject.” Rose II, 752 S.W.2d at 554.4
Rose II tells us we must presume the jury followed the instructions to disregard the effect of the parole laws. Rose II, 752 S.W.2d at 554. The presumption, it says, is rebuttable. Id. In this case, I believe the presumption the jury followed the instructions not to consider parole is rebutted for three reasons.
1. The Legislature intended the jury consider the parole laws.
In Rose II, the court told us to apply the rule 81(b)(2) harmless error test. Under that test, the State has the burden to prove the parole instruction played no role in the jury’s assessment of punishment. At the same time Rose II told us to apply the rule 81(b)(2) test, it also told us to presume the jury followed the statutory and curative instructions to disregard the parole instruction.
This presents us with a problem: How can we place the burden on the State and at the same time apply the presumption? If we require the State to prove beyond a reasonable doubt that the parole instruction did not contribute to the sentence, we must reverse almost all cases. If, however, we presume the jury did not consider the parole instructions, we must affirm almost all cases.
Rose II gives us vague instructions about applying conflicting tests to determine the harm of a fatally defective statute. We *486must start at the source of the problem: the statute. In order to attempt the harm analysis Rose II tells us to make, I think we must first understand how the statute was actually applied. That question, I believe, comes before we assume the jury followed the trial court’s instruction to disregard.
When we are not certain how a statute was actually applied, we look to see how the Legislature intended it to be applied. See Stanfield v. State 718 S.W.2d 734, 736 (Tex.Crim.App.1986) (op. on reh’g). To determine how the statute was applied, the Code Construction Act tells us to consider, among other things: the object sought to be attained; the circumstances under which the statute was enacted; the legislative history; the common law or former statutory provisions; the consequences of a particular construction; and the title, preamble, and emergency provisions. Tex.Gov’t Code Ann. sec. 311.023 (Vernon 1988).
We are required to interpret a statute to conform to the Legislature’s goal. See Alobaidi v. State, 433 S.W.2d 440, 442 (Tex.Crim.App.1968).5 We must, therefore, presume article 37.07, sec. 4 was applied to achieve the Legislature’s goal. To do that, we must look to the legislative history of article 37.07, sec. 4.
The revised bill analysis for S.B. 37, provided: 6
Background information:
There has been an outcry from public citizens serving as jurors that the sentences have been handed down have differentiated greatly from the sentences actually served. Jurors have indicated in some instances that they were recommending even longer sentences in order to compensate for the time which would be knocked off the sentence by the combination of good time credit and eligibility for parole. Problem(s) that the Bill Addresses:
Public citizens serving as jurors from across the state have indicated a need to be informed that the defendant may, but will not necessarily, be incarcerated for the full length of the sentence imposed; and the guidelines that are used to reduce the sentences through parole and good time credit.
How This Bill Will Solve the Problem^):
This legislation will allow the jury to be informed of these procedures.
In introducing S.B. 37, Senator Buster Brown explained:
We feel like based upon the public’s request, based upon a request from those working in the field of law enforcement, the judiciary, the prosecution, and citizen groups that are concerned about it, they’ve asked that we pass this bill. The University of Sam Houston Criminal Justice Division in their latest survey of Texans asked this question among the survey, and the results were 91% of people surveyed said they'd like for jurors to be told of parole. With that I would like to proceed to call the first witness.
Tex.Senate Criminal Justice Comm.Hearing [on tape], 69th Leg. (February 19, 1985).
The first witness, Rider Scott, an assistant district attorney representing the Dallas District Attorney’s Office, testified:
[Senate Bill 37] does provide some guidance for the juries, in that some jurors will assess a sentence of, let’s say, ten years, thinking that it will be truly assessing a five year sentence. Whereas other jurors may assess a sentence of ten years, with the misconception that he’s going to do every day of the ten years. [Senate Bill 37] provides some meaningful guidelines for jurors in exercising their discretion in setting a sentence for a defendant that is meaningful.
*487During Rider’s testimony, Senator McFarland asked Senator Brown the following question:
Senator McFarland: Assume eligibility is a third and jury is going to assess fifteen. And the jury, on your bill, says: well, that means he can get out in five, so let’s go ahead and make that if we wanted fifteen, 45 year sentence and we know he’s going to serve 15.
Senator Brown: A jury could do that. Senator McFarland: Do you not think that juries in almost every instance will do that?
To which Rider responded:
[I]f the [jurors] get back there with this type of instruction on a third eligibility, and they say: In 15 years he can be eligible for parole, and a third, let’s give him 45. Is that bad? Because, are you not then allowing the jury to express the intent of what they want that individual to do in the penitentiary for the crime that he has committed? That to me says that the jury has a meaningful understanding of not only the crime, and the defendant, but also the sentence that the individual is going to serve.... What I think this really does is bring some honesty to the sentencing process, so that the jury has an opportunity to know essentially what they’re doing.
[Emphasis added.] Tex.Senate Criminal Justice Comm.Hearing [on tape], 69th Leg. (February 19, 1935).
David Bires, president of the Texas Criminal Defense Lawyers Association, appeared before the Criminal Justice Committee and spoke against the bill:
There seem to be couple of assumptions that are operating here. One is that juries speculate on the operation of good time and parole law in assessing punishment. And I think we have to ask ourselves a question, and that is whether or not we are entitled to rely on the integrity of our trial jurors. These people are sworn in, they are sworn to follow the law of the case as given to them by the judge. They are, among other things, instructed that the matters of pardon and parole shall not be considered by them. That puts them on notice that there is such a thing as pardon and parole. I think the instruction that very succinctly tells them you shall not consider it, effectively does the job. They should not, and if they do, it can be a matter of misconduct.
⅜ ⅜ >⅛ ⅜ J* ⅞
What I am saying, is that [the jurors] are already admonished that they should not consider [parole]. Now, if they disregard that admonishment, that is a weakness in our system. But, to admonish them further that this is what it is and now that you know what it is, you won’t consider it as to this particular defendant, solves nothing. It only makes more complex that which they are going to disregard. My opinion is, that if you want people to be satisfied that the sentence they give is the sentence that will be served, get rid of good time and parole, if that is the object.
Tex.Senate Criminal Justice Comm.Hearing [on tape], 69th Leg. (February 26, 1985).
The last discussion before the Senate voted on S.B. 37, was:
Senator Brown: Members, I move to suspend all necessary rules to take up and consider out of its regular order, Senate Bill 37. Senate Bill 37 is a bill that deals with the subject of jury instruction on parole and good conduct time. Probably one of the, I guess, one of the most often called-for changes in our system by the public, those people who serve on juries in this state, is for the State to give them an opportunity to understand and have an idea about our parole system and good time system in this, state, so that they’ll have that knowledge before they go into the jury room and assess punishment. Yes, I yield.
Senator: Sir, what, does your bill actually allow the, charge the judge with the responsibility with telling the jury exactly when the defendant will be eligible for parole?
Senator Brown: No, Senator, it does not. In fact, it does admonish the jury on that subject and say they will not try *488to calculate the time upon which that person will be eligible for parole because that’s the province of the Board of Pardons and Parole, and not theirs.
Senator: As you know we’ve got a significant problem in prison over-crowding. Was there testimony indicating that this will have the effect of resulting in longer sentences, or any effect at all?
Senator Brown: Senator,, the testimony was, I guess, directed toward the fact that this will help bring about a more consistent setting of punishment, as opposed to increasing or decreasing, because they had testimony to show by district attorneys and by individuals that it will go both ways. In other words, there are juries who set punishment now, thinking there is some kind of a system, but they don’t know what it is, and so they will greatly enhance the punishment, give much more than they would ordinarily, in order to make sure that person does receive some time in the penitentiary. So, it’s our belief that this will bring about a more consistent setting of punishment, and the net result will not be a large increase in the amount of time that’s given.
Senator: That bill came to our committee, and in fact, several of the witnesses who were both prosecuting attorneys and defense attorneys, talked about that this bill may actually may serve to reduce the number of sentences that people receive, because many of the jurors know that there is a parole or probation system, and they decide that, well, this guy is going to get out in 2 years, so they give him 99 years. Do you recall that testimony?
Senator Brown: That’s correct.
Senator: Basically, what you are doing with this bill allowing the jury to assess the full — be fully knowledgeable about the effect of them verdict.
Senator Brown: That’s correct, Senator.
Debate on Tex. S.B. 37 on the Floor of the Senate, 69th Leg. (April 3, 1985) (on tape).
The Legislature adopted article 37.07, sec. 4 to respond to jurors’ unhappiness that parole laws reduced defendants’ sentences. The Legislature amended article 37.07 in 1985 to give the jurors correct information about the application of the parole laws in order to permit jurors to anticipate the executive branch’s exercise of clemency. The legislative history indicates the Legislature wanted the jurors to have accurate information about parole so when they ignored the instructions to disregard parole, the juries would consider accurate information about the parole laws instead of inaccurate information.7
If the Legislature intended the jury to invade the executive’s power, we must presume it was accomplished. Accordingly, we presume the jury disregarded trial court’s instruction to disregard the parole instruction. Once the presumption is rebutted, we look to see if under rule 81(b)(2) the State can prove the parole instruction did not contribute to the sentence. Only in those cases where the defendant received the lowest sentence permitted by law, can the State maintain its burden. I would, therefore, reverse all cases in which the defendant received more than the minimum punishment.
2. The Prominence of the Parole Issue in the Charge.
It is well settled that when the appellate courts review the charge, they do not review isolated statements in the charge but consider the charge as a whole. Jackson v. State, 591 S.W.2d 820, 825 (Tex.Crim.App.1979) (op. on reh’g). We must, therefore, look at the entire charge to determine the effect of the parole instruction.
*489The single most important subject in this charge was the parole instruction. No other subject received the same emphasis. The trial court submitted the charge in this case, like the charge in Rose, with both the statutory charge and the “curative” instruction. To illustrate the problem with the parole instruction in this case, I reproduce the charge from this case. All portions of the charge, except those dealing with the parole instructions, are summarized:
Paragraph 1: [One sentence introductory paragraph.]
Paragraph 2: [Two-sentence instruction on the statutory punishment for this crime.]
Paragraph 3: [One-sentence instruction to apply the punishment for this crime.] Paragraph 4: [Two-sentence instruction about defendant’s right to testify.] Paragraph 5: Under the law applicable to this case, the defendant, if sentenced to a term of imprisonment, may earn time off the sentence imposed through the award of good time. Prison authorities may award good conduct time to a prisoner who exhibits good behavior, diligence in carrying out prison work assignments, and attempts at rehabilitation. If a prisoner engages in misconduct, prison authorities may also take away all or part of any good conduct time earned by the prisoner. [Tex.Code Crim.P.Ann. art. 37.07, sec. 4(b)].
Paragraph 6: It is also possible that the length of time for which the defendant will be imprisoned might be reduced by the award of parole. [Tex.Code Crim.P. Ann. art. 37.07, sec. 4(b).]
Paragraph 7: Under the law applicable to this case, if the defendant is sentenced to a term of imprisonment, he will not become eligible for parole until the actual time served equals one-third of the sentence imposed or 20 years, whichever is less, without consideration of any good conduct time he may earn. If the defendant is sentenced to a term of less than six years, he must serve at least two years before he is eligible for parole. Eligibility for parole does not guarantee it will be granted. [Tex.Code Crim.P. Ann. art. 37.07, sec. 4(b).]
Paragraph 8: It cannot be accurately predicted how the parole law and good conduct time might be applied to this defendant because it depends on decisions made by the prison and parole authorities. [Tex.Code Crim.P.Ann. art. 37.07, sec. 4(b).]
Paragraph 9: You may consider the existence of the parole law and good conduct time. However, you may not consider the extent to which good conduct time may be awarded and forfeited by this particular defendant. You are not to consider the manner in which the parole law may be applied to this particular defendant by the authorities. [Tex.Code Crim.P.Ann. art. 37.07, sec. 4(b).]
Paragraph 10: You are further instructed that in fixing the defendant’s punishment, which you will show in your verdict, you may take into consideration all the facts shown by the evidence admitted before you in the full trial of this case and the law as submitted to you in this charge.
Paragraph 11: You are not to discuss among yourselves how long the defendant will be required to serve any sentence you decide to impose. Such matters come within the exclusive jurisdiction of the Board of Pardons and Parole and the Governor of the State of Texas and must not be considered by you. [This is the “curative instruction” added by some trial judges from the old charge. It is discussed in Rose II, 752 S.W.2d at 554.]
Paragraph 12: [One-sentence instruction that verdict must be unanimous.]
Paragraph 13: [One-sentence instruction that jury is exclusive judge of facts.]
Paragraph 14: [Four-sentence instruction that jury should only communicate with officer in charge of jury.]
Paragraph 15: [One-sentence instruction to retire to deliberate.]
The charge is two pages long; most of the charge is about parole. Of the 15 paragraphs in the charge, 6 paragraphs were about parole (paragraphs 5, 6, 7, 8, 9, *490and 11). The only other subject mentioned in more than one paragraph was the punishment, discussed in three sentences in two paragraphs (paragraphs 2 and 3). Of the 28 sentences in the charge, 14 sentences are about parole. In charging the jury, the trial court used the terms “parole” and “good time” 16 times. Considering the prominence of the parole instruction in this case, the presumption the jury did not consider parole as to this appellant is rebutted.
3. The Parole Instruction Was an Abstract Proposition of Law, Not Applied to the Facts of the Case.
Another rule we must apply in reviewing the charge is that it is error for the trial court to instruct the jury on an abstract proposition of law without applying it to the facts of the case. Newton v. State, 648 S.W.2d 693, 694 (Tex.Crim.App.1983). The instruction, to consider parole generally but not as to this defendant, was an abstract proposition of law that the trial court did not apply to the facts of this case.
The parole instruction told the jury to consider parole in this case but not as to this defendant. The jury, however, was assembled to make only one decision: what punishment is appropriate for this defendant? The jury was not brought together to ponder the philosophical consequences of the parole laws in this case. If the jury was to consider parole at all, it was only in their decision on punishment as to this defendant in this case. For this another reason, I would find the presumption that the jury followed the trial court’s instructions to disregard parole as to this appellant, was rebutted.
B. THE HEINOUS NATURE OF THE CRIME.
The Court in Rose II tells us to consider the heinous facts of the case in deciding whether the error in giving the parole instruction contributed to the punishment. Rose II, 152 S.W.2d at 553. In Rose II, the court affirmed the punishment because of the heinous nature of the crime.
The majority here held appellant participated in a violent crime from beginning to end.
C. APPELLANT’S CRIMINAL RECORD.
The second element Rose II tells us to consider was whether defendant had an criminal record. In Rose II, after listing defendant’s extensive criminal record, the court affirmed.
Appellant had no criminal record.
III. CONCLUSION
I would reverse and remand this case for a new hearing on punishment. First, in spite of the instructions to disregard, I believe the Legislature intended the jury to consider parole laws in assessing punishment. We must assume the Legislature was successful in designing a statute that informed the jury of the effect of the parole laws on their verdict. To apply the presumption that the jury disregarded parole is to presume the jury disregarded the Legislature’s intent in the statute. We cannot pretend the jury ignored what the Legislature intended it to consider. If we presume article 37.07, sec. 4 accomplished the Legislature’s goal, we should reverse all cases tried with the parole instruction except those cases where the defendant received the minimum sentence under the law.
Second, I would find the emphasis on the parole laws in this case assured the jury did not disregard parole. The only thing the instruction assured was that the jury would not discuss parole.
Third, because the parole instruction was an abstract proposition of law, not applied to the facts of the case, the case should be reversed.
Fourth, because appellant did not have a criminal record, I believe Rose II requires us to reverse.
In summary, I believe the jury probably considered the possibility appellant would receive parole, and the jury gave appellant *491a longer sentence because of the instructions on parole.
. Although the charge instructs on both good time and parole, I will refer to it as the "parole" instruction. Tex.Code Crim.P.Ann. 37.07, sec. 4, ch. 576, sec. 1, 1985 Tex.Gen.Laws 2195, amended by ch. 66, sec. 1, 1987 Tex.Gen.Laws 170, amended by ch. 1101, sec. 15, 1987 Tex.Gen. Laws 3765.
. See, for example, the decision of the San Antonio Court of Appeals in Olivarez v. State, 756 S.W.2d 113, 115 (Tex.App.—San Antonio 1988, no pet.), which takes issue with decisions of the Dallas Court of Appeals in Lancaster v. State, 754 S.W.2d 493 (Tex.App.—Dallas 1988, pet. refd), and Fort Worth Court of Appeals in Baker v. State, 752 S.W.2d 237 (Tex.App.—Fort Worth 1988, pet. refd). The San Antonio Court argues that the Fort Worth and Dallas Courts in post-Uose II opinions require proof of harm under the Almanza v. State, 686 $.W.2d 157 (Tex.Crim.App.1985) test, instead of the rule 81(b)(2) test. Olivarez, 756 S.W.2d at 115.
. Before the Legislature amended article 37.07, sec. 4 in 1985 to add the parole instruction, the trial court told the jury not to consider parole with this instruction:
You are instructed that you will not consider the amount of time the defendant will have to serve under any sentence you may assess, because such matters come within the exclusive jurisdiction of the Governor acting through the office of the Board of Pardons and Paroles.
After the effective date of the amendment to article 37.07, sec. 4, the courts submitted the instruction on parole as set out in the statute. When some defendants objected to the statutory charge on parole, some trial courts added the paragraph quoted above from the charge given the jury before the parole instruction of article 37.07. Many courts amended this paragraph to add: "and must not be considered by you.” This paragraph became known as the "curative" instruction.
. See this Court’s recent opinion in Washington v. State, 768 S.W.2d 497 (Tex.App.—Houston [1st Dist.] 1989, n.p.h.) (not yet reported), where Chief Justice Evans, writing for the majority, held the statutory instruction to disregard alone, was just as effective as the curative instruction to disregard. I agreed with the logic of that statement but dissented for other reasons.
. In construing an ambiguous statute, Alobaidi tells us to interpret it (1) to secure the Legislature’s intent, and (2) so that it will be constitutional. Alobaidi, 433 S.W.2d at 442. Even though Rose II already found the statute unconstitutional, we can still look to the Legislature’s intent to discover how the jury responded to the parole instruction.
. The amended bill, which this analysis summarized, required the judge to instruct the jury to consider parole in this case but not as to this defendant, as did the final version of the bill.
. Even now, as this case is decided, the Legislature is considering a constitutional amendment and a bill to inform juries about parole laws. The purpose statement for S.B. 34 states:
S.B. 34 reenacts Article 37.07, Section 4 of the Code of Criminal Procedure to allow for a jury charge which informs the jury of the operation of parole law and good time and which specifies the manner in which such matters can be considered by the jury. It is accompanied by S.J.R. 4, which would amend the constitution to explicitly empower the legislature to enact such instructions.
S.B. 34 provides the same instruction on parole, to consider parole in this case but not as to this defendant, as the one in article 37.07, sec. 4.